Voltage-controlled resistors for the PT100 / PT1000 or NTC (12k) simulation

[diemaiers]

Hello,

I am looking for a circuit with which I can simulate PT100 / PT1000 or NTCs with 12kOhm. I have to test temperature inputs of different CPC boards

I searched the forum but found nothing, the only ones are voltage-controlled current sinks their range but far away is from what I need.
I would have via DAC 12 / 16Bit a current sink build up the maximum 1mA at 3.3V can take up.
Has anyone of you, the more experience has a preface for me what I can begin.
Thank you very much
Tom

[kulky64]

Check TIPD120 from Texas Instruments here:
http://www.ti.com/tool/TIPD120
If you open main pdf document, on page 13 there is a simulation schematic for TINA-TI. Something similar can be probably done with LTSpice too.

[diemaiers]

I do not want take measure temperatures!
I want to simulate PT100 / PT100 NTC with a voltage-controlled resistance circuit or voltage-controlled current sink.

[2N3055]

So you need a PT100 Simulator/calibrator...

It can be done, but problem is to achieve required precision.. 
Speaking of:
1. so PT100, PT1000 and/or 12k NTC. I would recommend platinum sensor for linearity, NTC for better sensitivity..
2. What temperature range are we talking about?
3. Required precision

If you need this for production, you might need to buy a factory made, calibrated unit so you can guarantee  measurement results..

Take care ,
Sinisa

[diemaiers]

Ranges:
Pt100: 80 -120 Ohm /+-1%
PT1000: 800-1200 Ohm /+-1%
NTC:  2-380kOhm /+-1%
Operating Temp. for my board 20-70°C
The total Resistor Range for NTCs and PTs: 80 up to 400 kOhm
Over all max. 1mA / max. 3,3V
My board is for internal use only

Tank you

[Blue3]

For testing PT100 temperature inputs, you can use a PT100 RTD simulator. https://www.divize.com/signal-sources/a2041c.html The PT100 simulator is available in various temperature ranges. With the simulator, it will be easy to test your PT100 RTD temperature inputs. Unfortunately, this simulator is not suitable for calibration. For calibration, a set of precision resistors may be necessary. Hope it helps!

Regards,
Johann

[Zero999]

A voltage controlled resistor is not an easy thing to do and is virtually impossible to achieve a high accuracy. An ADC and digital potentiometer can be used, but the latter is not very accurate. An array of precision resistors switched by relays, controlled by an ADC is your best bet.

[b_force]

A voltage controlled resistor is not an easy thing to do and is virtually impossible to achieve a high accuracy. An ADC and digital potentiometer can be used, but the latter is not very accurate. An array of precision resistors switched by relays, controlled by an ADC is your best bet.

There is something called digipots (digital potentiometers).
When combining these with some analog resistors around the right value, you can have actually pretty precise control (with a limited range obviously).
When used over the full resistor value they lack accuracy.

My best bet would also be a current sink.
When used in parallel with some other (resistive) load you can have actually pretty accurate control.

Than being said, may I ask why you need such an high accuracy of reading temperatures?
The reason why I am asking, is because it is extremely difficult to have accurate temperature readings anyway.
From a physical point of view, not from an electrical point of view.
But even the latter ±0.5 degree C (Kelvin) is already very good and moving your probe a few mm or degrees has a more significant impact.
Not to talk about how tight it is connected and if any thermal grease is being used etc.


edit: another idea would be a motorized potentiometer.
But you have to pre calibrate it, or maybe put a little current sense circuit with it.

[Jay_Diddy_B]

Hi,

Take a look at this:

https://www.edn.com/multiplying-dac-makes-programmable-resistor/

You might just want to use a bunch of resistors and relays.

It depends if you how you need to test.

Regards,
Jay_Diddy_B

[Zero999]

A voltage controlled resistor is not an easy thing to do and is virtually impossible to achieve a high accuracy. An ADC and digital potentiometer can be used, but the latter is not very accurate. An array of precision resistors switched by relays, controlled by an ADC is your best bet.

There is something called digipots (digital potentiometers).
When combining these with some analog resistors around the right value, you can have actually pretty precise control (with a limited range obviously).
When used over the full resistor value they lack accuracy.

The problem is, he requires too lower resistance values for a digital potentiometer to work: 80 to 120 Ohm. It's possible to use an op-amp and MOSFET to effectively divide the resistance, but the accuracy problem still persists.

My best bet would also be a current sink.
When used in parallel with some other (resistive) load you can have actually pretty accurate control.

I don't see how that could work. Please post a schematic.

[b_force]

A voltage controlled resistor is not an easy thing to do and is virtually impossible to achieve a high accuracy. An ADC and digital potentiometer can be used, but the latter is not very accurate. An array of precision resistors switched by relays, controlled by an ADC is your best bet.

There is something called digipots (digital potentiometers).
When combining these with some analog resistors around the right value, you can have actually pretty precise control (with a limited range obviously).
When used over the full resistor value they lack accuracy.

The problem is, he requires too lower resistance values for a digital potentiometer to work: 80 to 120 Ohm. It's possible to use an op-amp and MOSFET to effectively divide the resistance, but the accuracy problem still persists.

My best bet would also be a current sink.
When used in parallel with some other (resistive) load you can have actually pretty accurate control.

I don't see how that could work. Please post a schematic.

In both cases (digipot as well as current), use a fixed resistor and put it parallel to that fixed resistor.
This gives you control over the current that's going through this circuit.
1kohm digipots do exist, so that is doable I think.

But once again, I am still confused why someone needs so much accuracy with temperature sensor to begin with.

[Jeroen3]

Do you really need a voltage controller resistor, or is a resistor multiplexer with some steps good enough?

[Zero999]

Looking at the specifications of digital potentiometers, the tolerance and temperature coefficients are very poor, nowhere near good enough to achieve the desired accuracy for this application. A chain of resistors bypassed with good old relays is probably the best solution.